the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Mar 2026
ESD Ideas: Reliable Adaptation Policies to Sea-Level Rise Require Incorporating Complexity in Economic Models
Abstract. With trillions of USD in assets facing climate-induced sea-level rise, oversimplified economic models might misinform responses. By capturing non-linearities in interconnected socio-economic and biogeophysical domains, and their local to global co-evolution, complexity science applied to sea-level rise uniquely enhances adaptation policies.
- Preprint
(1495 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 16 Apr 2026)
- CC1: 'Comment on egusphere-2026-792', Paul Pukite, 08 Mar 2026 reply
-
CC2: 'Comment on egusphere-2026-792', Judy Lawrence, 18 Mar 2026
reply
This ESD Idea is appealing and reflects the use of Agent Based modeling in a SLR context elsewhere. For example. there is a body of work in New Zealand in this context to reflect the concerns with current use of CBA in conditions of uncertainty and changing climate risk. Further research on agent based modeling in tandem with neural networks is underway. If the authors also have examples of how these approaches have been used it would enhance the utility of the idea. A collaboration would be nice.
Allison AEF, Lawrence JH, Stephens SA, Kwakkel JH, Singh SK, Blackett P and Stroombergen A (2024) Planning for wastewater infrastructure adaptation under deep uncertainty. Front. Clim. 6:1355446. doi: 10.3389/fclim.2024.1355446
Allison, A., Stephens, A., Blackett, P., Lawrence, J., Dickson, M., Matthews, Y. 2023. Simulating the Impacts of an Applied Dynamic Adaptive Pathways Plan Using an Agent-Based Model: A Tauranga City, New Zealand, Case Study. J. Mar. Sci. Eng. 2023, 11, x. https://doi.org/10.3390/xxxxx
Stroombergen, A and Lawrence, J (2022) A novel illustration of real options analysis to address the problem of probabilities under deep uncertainty and changing climate risk. Climate Risk Management 38, 2022, 100458 https://doi.org/10.1016/j.crm.2022.100458
Two improvements in this paper would help
1. The title could be improved . Suggest "Reliable sea level rise adaptation policies require complexity to be incorporated into economic models " or something shorter.
2. I like the Sankey diagram but the graphics in diagram B needs each community to be more clearly denoted. The Figure Panel descriptions need to be set out better as its a long read. Could this perhaps be incorporated into text in the main paper?
Citation: https://doi.org/10.5194/egusphere-2026-792-CC2 -
RC1: 'Comment on egusphere-2026-792', Robert Kopp, 18 Mar 2026
reply
This format is challenging to review. The article reads like an condensed excerpt from the motivation section of a funding proposal.
The first paragraph provides general background on sea-level rise; the second paragraph (uncited) presents well-known problems with classical theory-based climate impact economic models; the third and fourth pargraph critique classical approaches, particularly CGE models; the 5th-7th paragraph present ABMs as solutions; the 8th paragraph and the Figure 1 throw in a whole bunch of other solutions that are not going to be intelligible to someone not already versed in the literature; and a final sentence concludes.
The discussion of more classical economic approaches does not touch on approaches more innovative than traditional CGE (e.g., spatial integrated-assessment appraches approaches reviewed by Desmet and Rossi-Hansberg, 2026, https://doi.org/10.1093/jeg/lbaf049).
Agent-based modeling in climate policy analysis is not new (see Savin et al., 2023, https://doi.org/10.1002/wcc.811, for a review). Nor is it new to the sea-level rise context; a Web of Science search for "sea level" AND "agent-based" finds 54 papers.
The paper would thus benefit from a greater focus on what novel approach to ABM the authors are proposing, and how their proposed approach overcomes the barriers that have limited ABM uptake in the economic and policy analysis communities. I recognize that this is challenging to do in the space provided, but such discussion would greatly increase the value of the contribution.
Citation: https://doi.org/10.5194/egusphere-2026-792-RC1 -
RC2: 'Comment on egusphere-2026-792', Anonymous Referee #2, 23 Mar 2026
reply
This manuscript addresses an important topic: whether prevailing economic approaches to sea-level-rise adaptation omit key dynamics relevant to real decisions. The paper’s diagnostic critique is often persuasive. For example, it argues that existing approaches have important limitations regarding “bounded rationality, learning, risk perception,” “inequality and distributional impacts,” “indirect socio-economic losses,” and “adaptation processes over time.”
However, the manuscript is substantially stronger in diagnosing weaknesses in existing approaches than in substantiating the proposed alternatives. The abstract makes a very strong claim, stating that “complexity science applied to sea-level rise uniquely enhances adaptation policies.” Later, the paper concludes that “capturing interconnected nonlinear dynamics across domains and scales will improve assessments” while “facilitating reliable decision-making.” Those are consequential claims, but the body of the paper does not provide a comparative argument, empirical demonstration, or implementation-oriented assessment sufficient to support them at that level of confidence.
Major comments
1. The paper’s prescriptive case is underdeveloped relative to its diagnostic case.
The manuscript gives a compact but credible account of limitations in current CBA- and CGE-based approaches. It states, for example, that CBA can “omit the tail risk and the effect of extreme events,” and that CGE models may lead to “substantial underestimation of SLR damages and losses.” It also notes that local studies often ignore indirect damages, while macroeconomic assessments treat them only in aggregate. These are all legitimate concerns.
But when the manuscript turns to solutions, the argument becomes much thinner. The transition sentence is telling: “Here we present solutions for enhancing models of the economics of SLR, accompanied with visuals reflecting our opinion.” That phrasing signals that what follows is primarily a perspective, not a demonstrated comparative assessment. The manuscript then presents ABM, network analysis, and adaptive pathways as promising directions, but mostly at the level of possibility rather than demonstrated decision advantage.
That gap matters because the relevant question is not whether richer models can represent more mechanisms in principle. It is whether they improve decision support in practice, and under what conditions. The paper does not adequately engage with why simpler frameworks remain dominant in applied policy settings, including tractability, auditability, communicability, calibration burdens, and institutional familiarity. As written, the manuscript argues more for the promise of complexity methods than for their demonstrated superiority in adaptation decision-making.
2. Figure 1 is presented with the visual language of formal analysis but without a documented method.
This is, in my view, the clearest weakness of the manuscript. The paper says the visuals are “reflecting our opinion,” yet Figure 1 and its caption go well beyond a simple conceptual schematic. The caption states that “some issues are highly connected,” that “the size of nodes indicates the centrality,” and that node colours indicate “membership to communities of strongly interlinked items.” It also interprets substantive communities and identifies certain nodes as especially central.
That presentation strongly implies a formal analytic procedure. But the manuscript does not explain how issues and solutions were identified, how links were defined, whether weights were binary or continuous, how “centrality” was measured, or how “communities” were detected. Without that information, the figure is not reproducible and cannot be meaningfully evaluated. In its current form, it risks reading less like evidence and more like prior beliefs translated into the aesthetics of network analysis. At minimum, the paper would need to specify the source material, coding rules, graph construction procedure, weighting scheme, and community-detection method, or else present the figure more plainly as a qualitative conceptual diagram.
3. The manuscript does not establish that model complexity is the binding constraint on better adaptation policy.
The paper argues that current models “might misinform responses,” that some existing methods underestimate risk, and that more complex approaches would mitigate “risk underestimation and maladaptation.” That is plausible. But the key causal step is not demonstrated: the manuscript does not show that additional model complexity is the factor most likely to improve real policy outcomes, as opposed to other binding constraints such as institutional capacity, financing, governance fragmentation, or limited local data.
Relatedly, the paper moves too quickly from greater representational richness to greater policy usefulness. Those are not the same thing. In real decision contexts, transparency, interpretability, and communicability are often central to uptake. More complex models may improve realism on some dimensions while reducing usability or legitimacy on others. A paper advocating complexity-based approaches for policy support should address that tradeoff directly. As written, the manuscript assumes rather than demonstrates that more realistic models necessarily yield more reliable adaptation decisions.
Conclusion
This manuscript raises worthwhile concerns and could become a useful perspective piece. But in its current form it does not support its strongest claims. Its critique of existing frameworks is more convincing than its case for the superiority or practical policy relevance of the alternatives it recommends. Figure 1 is especially problematic because it adopts the appearance of formal analysis while the manuscript describes the visuals as “reflecting our opinion” and does not document the method needed to interpret the network structure it reports. As an Ideas piece, the manuscript need not provide a full comparative assessment. However, it does need to offer a well-founded and proportionate argument for why the proposed complexity-based approaches would improve adaptation decision-making in practice. In its current form, that case remains suggestive rather than adequately substantiated. For that reason, I do not recommend publication in its present form.
Citation: https://doi.org/10.5194/egusphere-2026-792-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 218 | 96 | 22 | 336 | 17 | 17 |
- HTML: 218
- PDF: 96
- XML: 22
- Total: 336
- BibTeX: 17
- EndNote: 17
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
What is first needed is a useful model of mean-sea-level variation as measured at the various coastal stations listed at PSMSL.org. This is the link between the relative simplicity of conventional daily tidal analysis and the large scale impact of ENSO. Once this is done, the discrimination and isolation of the increasing trend can be observed.
https:/github.com/pukpr/GEM-LTE